Cognitive Decline and Alzheimer's Disease in the Dallas Lifespan Brain Study

Episodic memory is a construct that measures how well individuals can store, maintain, and retrieve detailed information in long-term memory. Episodic memory will be a composite score using the Hopkins Verbal Learning test with 3 subcomponents (immediate recall: range 0-12, delayed recall: range 0-12, and recognition: range 0-24) and the immediate recall of the CANTAB Verbal Recognition Memory task (range 0-12). Scores from the tasks will be converted to Z-scores and averaged to form an episodic memory composite, and then this final value with be converted to a Z-score. A higher composite Z-score indicates better episodic memory, a Z-score of 0 represents the population mean, and all Z-scores have a standard deviation of 1. Values in this table represent this Episodic Memory Z-score.

Amyloid accumulation throughout the Dallas Lifespan Brain Study (6-9.8 years) will be measured with Florbetapir F18 and calculated as an Amyloid Standard Uptake Value ratio (SUVR) by normalizing regional counts to the whole cerebellum. Amyloid SUVR scores will be averaged across eight cortical regions spanning most of the cortex: dorsal lateral prefrontal cortex, orbitofrontal cortex, lateral parietal cortex, posterior cingulate cortex, anterior cingulate cortex, precuneus, lateral temporal cortex, and lateral occipital lobe. Amyloid SUVR scores observed in this study have ranged from 0.88 to 1.74, with higher scores being indicative of greater amyloid accumulation. Finally, annualized change scores for these amyloid SUVRs across the full study duration (6-9.8 years) will be calculated to determine the extent that the rate of amyloid accumulation relates to tau burden at the end of the study.

Speed of processing is a construct that measures how rapidly individuals can process information. To assess speed of processing, a composite score will be created using the Digit Comparison task and the Wechsler Adult Intelligence Scale (WAIS) Digit Symbol task. Observed DLBS raw scores range from 27 to 116 for Digit Comparison task and 20 to 90 for Digit Symbol task. Participants' raw scores are converted to Z-scores and averaged to form a speed of processing composite, and then this final value with be converted to a Z-score. A higher composite Z-score indicates better speed of processing, a Z-score of 0 represents the population mean, and all Z-scores have a standard deviation of 1. Values in this table represent this Speed of Processing Z-score.

The construct of reasoning measures an individual's ability to recognize novel patterns and the conceptual relationship among objects and effectively apply these patterns to solve similar problems. To assess reasoning, a composite score will be created using the Raven's Progressive Matrices task and Educational Testing Service (ETS) Letters Sets task. Observed DLBS raw scores range from 9 to 30 for Raven's Progressive Matrices task and from 0.5 to 30 for ETS Letters Sets task. Raven's Progress Matrices and ETS Letter Sets will be converted to Z-scores and averaged to form a reasoning composite, and then this final value with be converted to a Z-score. A higher composite Z-score indicates higher reasoning ability, a Z-score of 0 represents the population mean, and all Z-scores have a standard deviation of 1. Values in this table represent this Reasoning Z-score.

Working memory is a construct that measures the ability of individuals to simultaneously manipulate and store information. To assess working memory, a composite score will be created using the CANTAB Spatial Working Memory task (reverse scored) and the Wechsler Adult Intelligence Scale (WAIS-III) Letter Number Sequencing task. Observed DLBS raw scores range from 0 to 86 total errors for the CANTAB Spatial Working Memory task and 2 to 20 for the WAIS-III Letter Number Sequencing task. Participants' raw scores are converted to Z-scores and averaged to form a working memory composite, and then this final value with be converted to a Z-score. A higher working memory composite Z-score indicates better working memory, a Z-score of 0 represents the population mean, and all Z-scores have a standard deviation of 1. Values in this table represent this Working Memory Z-score.

Linear regressions between the AV-1451 SUVR in each of the temporal regions of interest with participant age will be calculated, and in additional analyses, non-linear effects of age will be examined via quadratic and growth modeling. Observed AV-1451 SUVR scores in humans have ranged from 0.5 to 2.5, with higher scores being indicative of greater tau accumulation. The investigators predict that tau accumulation will accelerate in old age, thus supporting a non-linear rate of deposition.

Regional cortical thickness will be estimated from previously acquired T1-weighted structural magnetic resonance imaging (MRl) scans using FreeSurfer (ver. 5.3), with surface parcellation manually edited when necessary by our team of experts. The regions selected for analyses of cortical thickness were the ROIs used to estimate temporal tau SUVR: inferior temporal gyrus, middle temporal gyrus, superior temporal gyrus, entorhinal cortex, parahippocampal gyrus, and fusiform gyrus. Observed cortical thickness in these regions range from 1.38 to 3.83 mm with higher scores indicating greater thickness.

Hippocampal volume will be estimated from previously acquired T1-weighted structural magnetic resonance imaging (MRl) scans using FreeSurfer (ver. 5.3), with surface parcellation manually edited when necessary by our team of experts. This region was selected for analysis as it was one of the ROIs used to estimate temporal tau SUVR. Observed DLBS hippocampal volume ranged from 1843 to 5342 mm^3 with higher scores indicating greater volume.

White matter integrity will be assessed using the estimated volume of white matter hypointensities from previously acquired T1-weighted structural magnetic resonance imaging (MRl) scans using FreeSurfer (ver. 5.3). Observed DLBS white matter hypointensities range from 796 to 35,037 mm^3 with higher scores indicating greater volume of hypointensities and reflecting worse white matter integrity.

For functional measures, blood oxygenation level dependent signal from contrasts of interest using selected ROls will be created. For the semantic judgment task (easy judgments - fixation), the investigators will focus on ROIs associated with processing meaning, including inferior frontal gyrus, precuneus, and middle temporal gyrus. Observed BOLD activation values (betas) ranged from -1.00 to 1.30 with higher values indicating greater activation.

Resting-state brain system segregation was computed on data collected from a separate resting-state scan using graph theory. System segregation is calculated as (Zw - Zb) / Zw, where Zw is the mean Fisher z-transformed r between nodes with the same system and Zb is the mean Fisher z-transformed r between nodes of one system to all nodes in other systems. Higher values indicate reduced node-node connectivity between systems relative to within-system connectivity. A score of 0 would reflect equal resting-state connectivity between nodes within the same functional system and between nodes of separate systems. Like all Z-scores, these scores typically range from -3 to 3. Higher scores are observed in younger adults than in older adults and may suggest a more youth-like brain, though higher scores are not objectively better. Systems were defined based on Power et al., (2011, Neuron) and more details on this measure are described in Chan et al. (2014, PNAS).